Apparatus for monitoring vehicle periphery

• US 5,729,216 A
• Filed: 04/22/1997
• Issued: 03/17/1998
• Est. Priority Date: 03/14/1994
• Status: Expired due to Fees

• Current Assignee: Yazaki Corporation
• Original Assignee: Yazaki Corporation
• Inventors: Ishikawa, Naoto, Fujinami, Kazutomo, Sasaki, Kazuyuki
• Primary Examiner(s): Mullen, Thomas
• Assistant Examiner(s): LIEU, JULIE BICHNGOC
• Application Number:
  US08/847,989
• Time in Patent Office: 329 Days
• Field of Search: 348/148, 348/149, 348/135, 348/139, 348/140, 348/142, 348/143, 348/159, 348/42, 348/43, 348/116, 348/113, 348/169, 340/937, 340/901, 340/903, 340/435, 382/104, 382/284, 382/294
• US Class Current: 340/937
• CPC Class Codes: B60Q 9/008 for anti-collision purposes
• CPC Class Codes: G01S 11/12 using electromagnetic waves other than radio waves
• CPC Class Codes: G01S 17/93 for anti-collision purposes

Abstract

A monitoring apparatus for monitoring the periphery of a motor vehicle in which patterns painted on the surface of a road are eliminated to shorten a time required for calculation to determine a position of a subject by processing image data generated by two cameras, whereby image data of the object having a height is derived by subtracting image data generated one camera from image data generating by the other camera after transferring either of image data to the other image coordinate by assuming that the transferring image data have a height of zero (0) for canceling road surface images and sounding an alarm whenever it is determined that the position of the object causes an obstacle of the moving vehicle.

Claims

1. A monitoring apparatus for monitoring the periphery of a vehicle, comprising:

first camera means which outputs first image data representing a view at a first position, and second camera means which outputs second image data to representing a view at a second position, said first camera means and said second camera means being mounted on the vehicle at said first and second positions, respectively, and spaced apart a predetermined distance;

a first memory and a second memory for respectively storing said first image data and said second image data;

road surface image elimination means, for preparing transformed first image data by transforming said first image data stored in said first memory to transformed first image data representing a view at said second position of images of said first image data with the assumption that all the images of said first image data are of no height, and for eliminating road surface images by computing a difference between said transformed first image data and said second image data, producing difference image data;

differentiation means for deriving differential image data differentiated in a horizontal direction from said second image data;

object edge detection means for detecting edge of an object by executing an AND operation of said difference image data and said differential image data;

object position calculation means for calculating a position of said object from said first and second image data stored in said first and second memories and said edges detected by said object edge detection means; and

alarm means for creating an alarm depending on the position calculated by said object position detection means.

2. A monitoring apparatus according to claim 1, further comprising distortion correction means for correcting image data stored in both of said first and second memories in consonance with the aberrations of the lenses of said first camera means and said second camera means.

3. A monitoring apparatus according to claim 1 wherein, upon computing a difference between said transformed first image data and said second image data, when a difference between corresponding picture element data of said transformed first image data and said second image data exceeds a specific value, a minimum difference value among differences that exist between one of said picture element data and a plurality of picture element data neighboring thereto is selected to be the difference between said two image data for that picture element data.

4. A monitoring apparatus according to claim 2 wherein, upon computing a difference between said transformed first image data and said second image data, when a difference between corresponding picture element data of said transformed first image data and said second image data exceeds a specific value, a minimum difference value among differences that exist between one of said picture element data and a plurality of picture element data neighboring thereto is selected to be the difference between said two image data for that picture element data.

5. A monitoring apparatus according to claim 1, 2, 3 or 4, wherein said differential image data differentiated in a horizontal direction is derived by utilizing a spatial filter.

6. A monitoring apparatus according to any one of claims 1 to 3 or 4, wherein said object position calculation is executed within a selected region.

7. A monitoring apparatus according to claim 1, wherein said object edge detection is implemented by executing AND operation after said difference image data and said differential image data are converted respectively to binary difference image data and binary differential image data at respective threshold levels.

8. A monitoring apparatus for monitoring the periphery of a vehicle, comprising:

first camera means which outputs first image data representing a view at a first position, and second camera means which outputs second image data representing a view at a second position, said first camera means and said second camera means being mounted on the vehicle at said first and second positions, respectively, and spaced apart a predetermined distance;

a first memory and a second memory for respectively storing said first image data and said second image data;

road surface image elimination means, for preparing transformed first image data by transforming said first image data stored in said first memory to transformed first image data representing a view at said second position of images of said first image data with the assumption that all the images of said first image data are of no height, and for eliminating road surface images by computing a difference between said transformed first image data and said second image data, producing difference image data;

object determination means for determining the presence of an object from said difference image data; and

alarm means for creating an alarm depending on the determination of said object position determination means.

9. A method for monitoring the periphery of a vehicle, comprising:

obtaining first image data representing a view at a first position, and obtaining second image data representing a view at a second position, said first and second positions being spaced apart a predetermined distance;

preparing transformed first image data by transforming said first image data to transformed first image data representing a view at said second position of images of said first image data with the assumption that all the images of said first image data are of no height;

eliminating road surface images by computing a difference between said transformed first image data and said second image data, producing difference image data;

deriving differential image data from said second image data;

detecting edges of an object by executing an AND operation of said difference image data and said differential image data;

calculating a position of said object from said first and second image data and said edges detected in said detecting step; and

creating an alarm in response to the position calculated in said calculating step.

10. The method of claim 9, wherein said first and second image data is obtained using first and second cameras, respectively, mounted on said vehicle at said first and second positions, respectively.

11. The method of claim 10, further comprising:

correcting said first and second image data as a function of aberrations, of said first and second cameras.

12. The method of claim 9 wherein;

said eliminating road surface step includes selecting, when a difference between corresponding picture element data of said transformed first image data and said second image data exceeds a specific value, a minimum difference value among differences that exist between one of said picture element data and a plurality of picture element data neighboring thereto, as the difference between said two image data for that picture element data.

13. A method for monitoring the periphery of a vehicle, comprising:

obtaining first image data representing a view at a first position, and second image data representing a view at a second position, said first and second positions being spaced apart a predetermined distance;

preparing transformed first image data by transforming said first image data to transformed first image data representing a view at said second position of images of said first image data with the assmuption that all the images of said first image data are of no height;

eliminating road surface images by computing a difference between said transformed first image data and said second image data, producing difference image data;

determining the presence of an object from said difference image data; and

creating an alarm in response to the determination of said object in said determining step.